1of20

The Segmentation Problem

Segmentation attempts to partition the pixels of an image into groups that strongly correlate with the objects in an image

Typically the first step in any automated computer vision application

2of20

Bahadir K. Gunturk EE 7730 - Image Analysis I

2

Image Segmentation

Group similar components (such as, pixels in an image, image frames in a video) to obtain a compact representation.

Applications: Finding tumors, veins, etc. in medical images, finding targets in satellite/aerial images, finding people in surveillance images, summarizing video, etc.

Methods: Thresholding, region growing, k-means.

3of20

Bahadir K. Gunturk EE 7730 - Image Analysis I

3

Image Segmentation

Segmentation algorithms for monochrome images generally are based on one of two basic properties of gray-scale values:

Discontinuity The approach is to partition an image based on abrupt changes

in gray-scale levels. The principal areas of interest within this category are detection

of isolated points, lines, and edges in an image. Similarity

The principal approaches in this category are based on thresholding, region growing, and region splitting/merging.

4of20

Segmentation ExamplesIm

ag

es

take

n f

rom

Go

nza

lez

& W

oo

ds,

Dig

ital I

ma

ge

Pro

cess

ing

(2

00

2)

5of20

Detection Of Discontinuities

There are three basic types of grey level discontinuities that we tend to look for in digital images:

– Points– Lines– Edges

We typically find discontinuities using masks and correlation

6of20

Point Detection

Point detection can be achieved simply using the mask below:

Points are detected at those pixels in the subsequent filtered image that are above a set threshold

Ima

ge

s ta

ken

fro

m G

on

zale

z &

Wo

od

s, D

igita

l Im

ag

e P

roce

ssin

g (

20

02

)

7of20

Point Detection (cont…)Im

ag

es

take

n f

rom

Go

nza

lez

& W

oo

ds,

Dig

ital I

ma

ge

Pro

cess

ing

(2

00

2)

X-ray image of a turbine blade

Result of point detection

Result of thresholding

8of20

Line Detection

The next level of complexity is to try to detect lines

The masks below will extract lines that are one pixel thick and running in a particular direction

Ima

ge

s ta

ken

fro

m G

on

zale

z &

Wo

od

s, D

igita

l Im

ag

e P

roce

ssin

g (

20

02

)

9of20

Line Detection (cont…)Im

ag

es

take

n f

rom

Go

nza

lez

& W

oo

ds,

Dig

ital I

ma

ge

Pro

cess

ing

(2

00

2)

Binary image of a wire bond mask

After processing

with -45° line detector

Result of thresholding filtering result

10of20

Edge Detection

An edge is a set of connected pixels that lie on the boundary between two regions

Ima

ge

s ta

ken

fro

m G

on

zale

z &

Wo

od

s, D

igita

l Im

ag

e P

roce

ssin

g (

20

02

)

11of20

Edges & Derivatives

We have already spokenabout how derivatives are used to find discontinuities

1st derivative tells us where an edge is

2nd derivative canbe used to show edge direction

Ima

ge

s ta

ken

fro

m G

on

zale

z &

Wo

od

s, D

igita

l Im

ag

e P

roce

ssin

g (

20

02

)

12of20

Derivatives & Noise

Derivative based edge detectors are extremely sensitive to noise

We need to keep this in mind

Ima

ge

s ta

ken

fro

m G

on

zale

z &

Wo

od

s, D

igita

l Im

ag

e P

roce

ssin

g (

20

02

)

13of20

Common Edge Detectors

Given a 3*3 region of an image the following edge detection filters can be used

Ima

ge

s ta

ken

fro

m G

on

zale

z &

Wo

od

s, D

igita

l Im

ag

e P

roce

ssin

g (

20

02

)

14of20

Edge Detection ExampleIm

ag

es

take

n f

rom

Go

nza

lez

& W

oo

ds,

Dig

ital I

ma

ge

Pro

cess

ing

(2

00

2)

Original Image Horizontal Gradient Component

Vertical Gradient Component Combined Edge Image

15of20

Edge Detection Problems

Often, problems arise in edge detection in that there are is too much detail

For example, the brickwork in the previous example

One way to overcome this is to smooth images prior to edge detection

16of20

Edge Detection Example With Smoothing

Ima

ge

s ta

ken

fro

m G

on

zale

z &

Wo

od

s, D

igita

l Im

ag

e P

roce

ssin

g (

20

02

)

Original Image Horizontal Gradient Component

Vertical Gradient Component Combined Edge Image

17of20

Laplacian Edge Detection

We encountered the 2nd-order derivative based Laplacian filter already

The Laplacian is typically not used by itself as it is too sensitive to noiseUsually then used for edge detection the Laplacian is combined with a smoothing Gaussian filter

Ima

ge

s ta

ken

fro

m G

on

zale

z &

Wo

od

s, D

igita

l Im

ag

e P

roce

ssin

g (

20

02

)

18of20

Laplacian Of Gaussian

The Laplacian of Gaussian (or Mexican hat) filter uses the Gaussian for noise removal and the Laplacian for edge detection

Ima

ge

s ta

ken

fro

m G

on

zale

z &

Wo

od

s, D

igita

l Im

ag

e P

roce

ssin

g (

20

02

)

19of20

Laplacian Of Gaussian ExampleIm

ag

es

take

n f

rom

Go

nza

lez

& W

oo

ds,

Dig

ital I

ma

ge

Pro

cess

ing

(2

00

2)

20of20

Edge Linking and Boundary Detection

• Local Processing

• Global Processing

• Graph-Theoretic Techniques

21of20

Local Processing

• Edge Linking using gradient and gradient direction.

22of20

23of20

24of20

Graph-Theoretic Techniques

25of20

26of20

27of20

28of20

29of20

30of20

31of20